Week 1: Neural Network Basics

Objective

Build intuition for a single neuron: inputs go in, weights scale them, a bias shifts the result, and an activation function shapes the final output.

Required Videos

• 3Blue1Brown: “But what is a Neural Network?”
• Optional: rewatch the first 15 minutes after running the code

Tasks

1. Run python week1_neuron.py.
2. Change the weights and predict how the output will change before rerunning.
3. Try at least two different inputs and compare the raw score vs. activated score.
4. Explain in your own words why the activation function changes the behavior.

Deliverables

• A screenshot or copied output from one successful run
• One modified version of the weights or inputs
• A short written explanation of how the neuron makes decisions

Checkpoint Questions

• What does a positive weight do?
• What does a negative weight do?
• What is the purpose of the bias term?
• Why is the activated output different from the raw weighted sum?

Stretch Idea

Replace the tanh activation with a sigmoid function and compare the output range.